[Simh] Problem with reading tape with PDP-11 SIMH

[Timothe Litt]

On 21-May-15 15:44, Alan Frisbie wrote:
> Before I dig too deeply into this problem, I thought I would ask
> if anyone else has seen it, or has some insight.
>
> I am running a relatively recent (mid-April) version of the
> PDP-11 simulator on my Alphastation XP1000 with VMS v8.4.
> The simulated PDP-11 is a copy of the physical 11/73 system
> I had running RSX-11M v3.0 back in 1986.
>
> So far, all simple things work just fine.
>
> For instance, I can ATTACH a .TAP tape container file to SIMH
> TS0, and read it (as RSX device MS0:) OK using FLX.   I can
> also use DMP and other utilities to read the "tape".
>
> The problem comes when I try to run TPC on the RSX system to
> create a container file.   Since I was the TPC maintainer back
> then, I am confident that it works correctly (at least for this
> purpose).   What appears to happen is that the first 16 or so
> reads of the tape complete OK, and the data gets written to
> the new container file on the disk.   However, nothing happens
> after that.
>
> If I abort TPC and dump the next record on the "tape", it appears
> to be about 16 records in from the start.   (I'll verify the
> exact numbers next time).
>
> >From this, it looks like SIMH may not be handling the read
> requests after the first burst, or may not be properly giving
> completion interrupts to the PDP-11.   Remember, TPC was written
> to move a tape drive as fast as possible, so it will stress any
> emulator.
>
> I hate to blame SIMH so quickly, particularly with so few solid
> facts, but this system (at least on physical hardware) worked
> perfectly.
>
> Does anyone have any ideas or suggestions before I start a deep
> dive into the code?
>
> Thanks,
> Alan Frisbie
> _______________________________________________
> Simh mailing list
> Simh at trailing-edge.com
> http://mailman.trailing-edge.com/mailman/listinfo/simh
I don't have an answer, but as it happens I'm deep into simh tapes at the
moment (for reasons that will become apparent when I dig myself out.)

[I'm into the tape formats/handling, not the device emulations, but that
 has required some digging.]

Some questions to get you started:

You may be right that completion interrupts are dropped.  Does this happen
only on the TS11 emulation?  Have you tried other drive types?

Do you build the simulator with the asynchronous (threaded) tape IO?
If so, do you see the failures if you don't?  (I believe that only some
of the
drive emulations use async io; I'm not current on which do and which don't.)

Does your input tape have any of the unusual cases:
  records marked with data errors
  erase gaps
  records written over erase gaps
? 

Are the record sizes fixed?  If you create a tape with different record
sizes,
does the symptom change?  Is 16 the number of buffers that TPC allocates?

I have in progress (but not ready to release) a utility that dumps a .tap
file & will report on its structure.  If a dump of your tape would help,
contact me off-line.


-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 4942 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://mailman.trailing-edge.com/pipermail/simh/attachments/20150521/bf90bd95/attachment.bin>